The authors of the paper argue that algorithms no longer simply parse scientific literature but are capable of synthesizing new, non-obvious biomedical hypotheses. Moreover, these AI-generated ideas have already moved to the physical testing stage: they are being tested on organoids, animal models, and even in early-phase clinical trials. This radically accelerates the Drug Discovery cycle, relieving the cognitive load on research teams and eliminating human bias when searching for correlations in molecular biology.
Source: Nature Medicine
ScienceMedicineResearchNatureDrug Discovery